What is work?
Here is a formula for work:
$$t_w = t_{exec} + \frac{n_h(n_h - 1)}{2} \cdot \frac{\tau}{\phi}$$
Most people have never seen this written down. Most people have never even thought to ask the question. But if you run a business and your time is a real input, this equation is the most important thing you are not measuring.
Let's unpack it.
First, a Scope Check: This Isn't About Passive Investment
Before going further, let's be precise about what kind of business we are talking about.
If you put $100k into an index fund and get back $130k, ROI is the right metric. Your time was not a meaningful input. The equation is clean:
$$\frac{m_{out}}{m_{in}} > 1$$
But that is not what most founders, indie hackers, or small business operators are doing. They are inside the business. They are building it, running it, fixing it, selling it. Their time is not a side effect. It is a primary input, as real as the capital.
And yet they still use ROI as their scoreboard.
This is where the deception starts. ROI looks at money in vs. money out, and says nothing about the hours you burned generating that return. A business doing $500k/year with you working 80-hour weeks scores identically to one doing $500k/year while you work 10-hour weeks. Same ROI. Completely different lives.
ROI is the right tool for passive capital. It is the wrong tool for operational businesses where your time is on the table.
Everything that follows is for the second kind: the business you are in, not just invested in.
What Even Is "Work"?
Before we can measure work, we need to define it.
The naive answer is: whatever you do for the business. But that breaks down fast.
Is reading about a competitor over breakfast work? Is a shower thought about a product feature work? What about the Sunday anxiety about Monday? What about a coffee with someone who might become a partner?
A more useful definition: work is time the business has a claim on.
The key criterion is not activity. It is obligation. Work is any block of time where saying "not now" has a consequence. Where you are not free to redirect your attention without cost.
Under this definition, a 9am call you must attend is obviously work. But so is lying awake at 2am because a server might be down. The activity is different. The claim is the same.
This matters because it reveals a hidden category most founders never account for: mental occupancy. The business living rent-free in your head during nominally free time. The background thread that never quite terminates.
A founder who logs 4 hours of work per day but thinks about the business for 12 is not running a lifestyle business. They are running a mental-overhead business with occasional gaps.
Three Zones, Not Two
Rather than a binary work/free split, the model becomes more honest with three zones:
- $t_w$ -- obligated time: the business has a hard claim. You must be present, responsive, or accountable.
- $t_g$ -- chosen engagement: you are voluntarily working on the business, but you could stop. It does not feel like cost, but it is still consuming time.
- $t_f$ -- genuinely free time: the business has no claim, mental or otherwise. You are fully off.
The $t_g$ zone is interesting. A founder who loves building might spend Saturday afternoon coding because they want to, not because they have to. This is not the same as $t_w$. But it is also not fully $t_f$: it is time that could have gone elsewhere, voluntarily redirected to the business.
What a lifestyle business is really optimizing for is $t_f$: the unoccupied fraction of your life. Not fewer hours logged, but fewer hours owned.
$$t_f = T - t_w - t_g$$
Most entrepreneurs fail not because they work too much in the obvious sense, but because $t_g$ quietly expands to fill every gap, and mental occupancy turns nominally free time into something that looks free but is not.
The Work Equation
Now back to the formula we opened with.
Work has exactly two components.
The first is execution:
$$t_{exec}$$
This is time you spend on productive output: writing code, making decisions, creating things. It is the part of work most people think of when they think of work.
The second is coordination:
$$t_{coord}(n_h, \phi) = \frac{n_h(n_h - 1)}{2} \cdot \frac{\tau}{\phi}$$
Where:
- $n_h$ is the number of humans in your operation
- $\tau$ is the average time cost of each communication channel (meetings, clarifications, corrections, check-ins)
- $\phi \in (0, 1]$ is a fidelity coefficient representing how accurately your intent is understood and executed
Put together:
$$\boxed{t_w = t_{exec} + \frac{n_h(n_h - 1)}{2} \cdot \frac{\tau}{\phi}}$$
This is the equation for work. The first term you control directly. The second term controls you, unless you design against it.
The Quadratic Coordination Tax
The term $\frac{n_h(n_h - 1)}{2}$ is the number of communication channels between $n_h$ people. It comes from combinatorics and is the same result underlying Brooks's Law in software engineering.
With 2 people: 1 channel. With 5 people: 10 channels. With 10 people: 45 channels.
Each channel carries a recurring time cost. This is not a linear relationship. It is quadratic. Adding one person does not add one unit of coordination overhead. It adds $n_h$ new channels all at once.
The implication is counterintuitive but mathematically unavoidable: adding people can increase the founder's working time even as they do less actual work. You are delegating $t_{exec}$ but importing $t_{coord}$ at a faster rate.
The Understanding Gap: Fidelity
There is a second problem with humans that the channel count alone understates.
When you communicate intent to another person, something is lost at every step:
- What you mean is not what you say
- What you say is not what they hear
- What they hear is not what they do
This is not incompetence. It is the nature of human communication. Language is lossy. Context is assumed. Priorities are interpreted differently by different people.
The fidelity coefficient $\phi$ captures this. Lower $\phi$ means more correction loops: more re-explaining, more reviewing, more fixing the gap between what you wanted and what arrived. It sits in the denominator of the coordination term, which means even small drops in fidelity have large effects on total working time.
A team of 5 with $\phi = 0.5$ generates twice the coordination overhead of a team of 5 with $\phi = 1.0$. Same headcount, same channels, twice the tax.
Systems vs. Humans: The Core Asymmetry
Now we can see what makes systems, software, automation, and AI agents structurally different from human workers.
A machine, once built, has:
- $\phi \approx 1$: it does exactly what it is told
- Near-zero ongoing coordination cost: it does not need motivation, context, or check-ins
- A one-time setup cost $t_{setup}$ that is then amortized across unlimited future executions
A human incurs $t_{coord}$ and $\phi < 1$ every single day, forever.
So we can split the workforce into two types:
- $n_h$ = human workers, with fidelity $\phi < 1$ and quadratic coordination drag
- $n_s$ = systems/machines, with $\phi \approx 1$ and setup cost only
The full coordination term becomes:
$$t_{coord} = \underbrace{\frac{n_h(n_h - 1)}{2} \cdot \frac{\tau}{\phi}}_{\text{human overhead}} + \underbrace{t_{setup}}_{\text{one-time machine cost}}$$
Pay once for setup. Then coordination approaches zero.
The Lifestyle Return
With the work equation defined, we can now write the metric that actually matters for an operational business: the Lifestyle Return.
$$LR = \frac{m_{out}}{m_{in}} \times \frac{t_f}{t_w}$$
Where $t_f = T - t_w - t_g$ is genuinely free time, not just non-working time.
A standard high-revenue business might have a financial return of 10x but a time ratio of 0.1. The Lifestyle Return is 1. You broke even on life.
The lifestyle business condition is $LR \gg 1$: both financial and temporal ratios above 1, and their product as high as possible.
Substituting the full work equation:
$$\boxed{LR = \frac{m_{out}}{m_{in}} \times \frac{t_f}{t_{exec} + \dfrac{n_h(n_h-1)}{2} \cdot \dfrac{\tau}{\phi}}}$$
Every variable is now visible. Every lever is explicit.
Does This Model Connect to Anything Known?
Yes, in several places.
Transaction Cost Economics (Coase, 1937; Williamson, 1975)
Ronald Coase asked a deceptively simple question in "The Nature of the Firm": if markets are efficient, why do firms exist at all? His answer was transaction costs. Firms emerge because they have lower transaction costs than the market. But a firm cannot endlessly expand because it also has internal coordination costs: administrative costs, coordinating costs, and the cost of preventing opportunistic behaviour among employees.
Our $t_{coord}$ term is precisely this internal transaction cost, now expressed as a function of headcount and fidelity rather than left qualitative. Coase observed that "diminishing returns to management means rising costs per task when a firm expands the range of productive activities coordinated by its managers." Our quadratic term formalizes exactly that observation.
Agency Theory (Jensen and Meckling, 1976)
The fidelity coefficient $\phi$ maps directly onto the principal-agent problem. Agency theory recognizes conflicts of interest between different economic actors and deals with the problems resulting from the principal-agent relationship, such as adverse selection and moral hazards. When $\phi < 1$, some fraction of the principal's intent is lost or distorted by the agent. The correction loops that result are the agency costs expressed in time rather than money.
Brooks's Law (Fred Brooks, "The Mythical Man-Month", 1975)
The $\frac{n(n-1)}{2}$ channel count is the direct mathematical foundation of Brooks's famous observation that adding manpower to a late software project makes it later. We borrow the combinatorial structure and generalize it beyond software to any operational business with human workers.
Tim Ferriss, "The 4-Hour Workweek" (2007)
Ferriss introduced the concept of "relative income," defined as hourly income, in contrast to "absolute income," which is total salary without regard to time. Relative income can be increased by increasing total income for the same hours, getting the same income for fewer hours, or some combination of both. The Lifestyle Return is a formalization of this intuition: relative income is the ratio $m_{out} / t_w$, and we add the temporal dimension explicitly.
Where our model goes beyond Ferriss is in decomposing why $t_w$ is large. He prescribes working less. We show what makes working less structurally difficult: the quadratic coordination term that grows with every human added to the operation.
What the Equation Tells You to Do
The levers for maximizing $LR$ are now explicit.
1. Minimize $n_h$ aggressively Because the drag is quadratic, removing one human has outsized effect. You are not just saving that person's salary. You are eliminating all the channels they create.
2. Maximize $\phi$ Where humans are necessary, invest in making them more machine-like: clear documentation, SOPs, well-defined scopes, trained judgment. Each improvement in $\phi$ reduces correction overhead multiplicatively.
3. Shift from $n_h$ to $n_s$ Pay the setup cost once. Then coordination approaches zero. This is the structural advantage of SaaS, automation, and AI agents: the marginal time cost of each additional unit of output trends toward zero.
4. Keep $t_{exec}$ small This is the classic lifestyle design move. But it is only a piece of the puzzle. If $t_{coord}$ is large enough, reducing $t_{exec}$ barely moves the needle.
The Real Insight
The enemy of a lifestyle business is not work. It is management.
A business with 10 employees doing the same revenue as a business with 2 employees and good systems is structurally worse: not because of the payroll, but because of the hidden $n^2$ time tax imposed on the founder.
The lifestyle entrepreneur's real job is not to work less. It is to architect systems where the work happens without them, and where the understanding gap between intent and execution has been closed by design, not by constant supervision.
A machine does not need motivation. It does not misinterpret the brief. It does not require a 1:1.
It just runs.
A Note on What This Does Not Capture
This model is deliberately simplified. There are real things it does not account for:
- Mental occupancy is real but hard to measure: $t_g$ expands to fill gaps in ways that are easy to rationalize as freedom
- The quality of $t_f$ matters, not just its quantity: burned-out free time is not free
- Humans create things systems cannot: relationships, creativity, trust, serendipity
- High $n_h$ can be worth it in early phases when learning velocity matters more than time efficiency
- $m_{out}$ itself often depends on human inputs that resist automation
The equation is a thinking tool, not a prescription. Use it to audit where your time is actually going, and whether the business you have built is returning it to you.
The standard business question is: does this make money? The lifestyle business question is: does this make money, and does it give me back my time?
Most entrepreneurs only ask the first one.
Mike Rubini is the founder of FlatNine, a bootstrapped SaaS portfolio. He writes about building businesses that compound in silence.